LENGTH ESTIMATION OF DIGIT STRINGS USING A NEURAL-NETWORK WITH STRUCTURE-BASED FEATURES

Accurate length estimation is very helpful for the successful segmenta tion and recognition of connected digit strings, in particular, for an off-line recognition system. However, little work has been done in th is area due to the difficulties involved. A length estimation approach is presented as a part of our automatic off-line digit recognition sy stem. The kernel of our approach is a neural network estimator with a set of structure-based features as the inputs. The system outputs are a set of fuzzy membership grades reflecting the degrees of an input di git string of having different lengths. Experimental results on Nation al Institute of Standards and Technology (NIST) Special Database 3 and other derived digit strings shows that our approach can achieve an ab out 99.4% correct estimation if the best two estimations are considere d. (C) 1998 SPIE and IS&T. [S1017-9909(98)00901-5].